This invention can be used with the invention described in my co-pending U.S. patent application Ser. No. 09/920,437 filed Aug. 1, 2001, now U.S. Pat. No. 6,659,016.
Auto rack rail road cars are used to transport automobiles. Typically, auto-rack rail road cars are loaded in the “circus loading” manner, by driving vehicles into the cars from one end, and securing them in places with chocks, chains or straps. When the trip is completed, the chocks are removed, and the cars are driven out.
Automobiles are a high value, relatively fragile type of lading. Damage due to dynamic loading in the railcar may tend to arise principally in two ways. First, there are the longitudinal input loads transmitted through the draft gear due to train line action or shunting. Second, there are vertical, rocking and transverse dynamic responses of the rail road car to track perturbations as transmitted through the rail car suspension. It would be desirable to improve ride quality to lessen the chance of damage occurring.
In the context of longitudinal train line action, damage most often occurs from two sources (a) slack run-in and run out; (b) humping or flat switching. Rail road car draft gear have been designed against slack run-out and slack run-in during train operation, and also against the impact as cars are coupled together. Historically, common types of draft gear, such as that complying with, for example, AAR specification M-901-G, have been rated to withstand an impact at 5 m.p.h. (8 km/h) at a coupler force of 500,000 Lbs. (roughly 2.2×106 N). Typically, these draft gear have a travel of 2¾ to 3¼ inches in buff before reaching the 500,000 Lbs. load, and before “going solid”. The term “going solid” refers to the point at which the draft gear exhibits a steep increase in resistance to further displacement. If the impact is large enough to make the draft gear “go solid” then the force transmitted, and the corresponding acceleration imposed on the lading, increases sharply. While this may be acceptable for ores, coal or grain, it is undesirably severe for more sensitive lading, such as automobiles or auto parts, rolls of paper, fresh fruit and vegetables and other high value consumer goods such as household appliances or electronic equipment. Consequently, from the relatively early days of the automobile industry there has been a history of development of longer travel draft gear to provide lading protection for relatively high value, low density lading, in particular automobiles and auto parts, but also farm machinery, or tractors, or highway trailers.
Historically, the need for slack was related, at least in part, to the difficulty of using a steam locomotive to “lift” (that is, move from a standing start) a long string of rail road cars with journal bearings, particularly in cold weather. For practical purposes, presently available diesel-electric locomotives are capable of lifting a unit train of one type of cars having little or no slack. Given the availability of locomotives that develop continuous high torque from a standing start, it is possible to re-examine the issue of slack action from basic principles. By eliminating, or reducing, the accumulation of slack, the use of short travel buff gear may tend to reduce the relative longitudinal motion between adjacent rail road cars, and may tend to reduce the associated velocity differentials and accelerations between cars. The use of short travel, or ultra-short travel, buff gear also has the advantage of eliminating the need for relatively expensive, and relatively complicated EOCC units, and the fittings required to accommodate them.
In terms of dynamic response through the trucks, there are a number of loading conditions to consider. First, there is a direct vertical response in the “vertical bounce” condition. This may typically arise when there is a track perturbation in both rails at the same point, such as at a level crossing or at a bridge or tunnel entrance where there may be a relatively sharp discontinuity in track stiffness. A second “rocking” loading condition occurs when there are alternating track perturbations, typically such as used formerly to occur with staggered spacing of 39 ft rails. This phenomenon is less frequent given the widespread use of continuously welded rails, and the generally lower speeds, and hence lower dynamic forces, used for the remaining non-welded track. A third loading condition arises from elevational changes between the tracks, such as when entering curves in which case a truck may have a tendency to warp. A fourth loading condition arises from truck “hunting”, typically at higher speeds, where the truck oscillates transversely between the rails. During hunting, the trucks tend most often to deform in a parallelogram manner. Fifth, lateral perturbations in the rails sometimes arise where the rails widen or narrow slightly, or one rail is more worn than another, and so on.
There are both geometric and historic factors to consider related to these loading conditions. One historic factor is the near universal usage of the three-piece style of freight car truck in North America. While other types of truck are known, the three piece truck is overwhelmingly dominant in freight service in North America. The three piece truck relies on a primary suspension in the form of a set of springs trapped in a “basket” between the truck bolster and the side frames. For wheel load equalisation, a three piece truck uses one set of springs, and the side frames pivot about the truck bolster ends in a manner like a walking beam. The 1980 Car & Locomotive Cyclopedia, states at page 669 that the three piece truck offers “interchangeability, structural reliability and low first cost but does so at the price of mediocre ride quality and high cost in terms of car and track maintenance”. It would be desirable to retain many or all of these advantages while providing improved ride quality.
In terms of rail road car truck suspension loading regimes, the first consideration is the natural frequency of the vertical bounce response. The static deflection from light car (empty) to maximum laded gross weight (full) of a rail car at the coupler tends to be typically about 2 inches. In addition, rail road car suspensions have a dynamic range in operation, including a reserve travel allowance.
In typical historical use, springs were chosen to suit the deflection under load of a full coal car, or a full grain car, or fully loaded general purpose flat car. In each case, the design lading tended to be very heavy relative to the rail car weight. For example, the live load for a 286,000 lbs. car may be of the order of five times the weight of the dead sprung load (i.e., the weight of the car, including truck bolsters but less side frames, axles and wheels). Further, in these instances, the lading may not be particularly sensitive to abusive handling. That is, neither coal nor grain tends to be badly damaged by poor ride quality. As a result, these cars tend to have very stiff suspensions, with a dominant natural frequency in vertical bounce mode of about 2 Hz. when loaded, and about 4 to 6 Hz. when empty. Historically, much effort has been devoted to making freight cars light for at least two reasons. First, the weight to be back hauled empty is kept low, reducing the fuel cost of the backhaul. Second, as the ratio of lading to car weight increases, a higher proportion of hauling effort goes into hauling lading, rather than hauling the railcar.
By contrast, an autorack car, or other type of car for carrying relatively high value, low density lading such as auto parts, electronic consumer goods, or white goods more generally, has the opposite loading profile. A two unit articulated autorack car may have a light car (i.e., empty) weight of 165,000 lbs., and a lading weight when fully loaded of only 35-40,000 lbs., per car body unit. That is, not only may the weight of the lading be less than the sprung weight of the rail road car unit, it may be less than 40% of the car weight. The lading typically has a high, or very high, ratio of value to weight. Unlike coal or grain, automobiles are relatively fragile, and hence more sensitive to a gentle (or a not so gentle) ride. As a relatively fragile, high value, high revenue form of lading, it may be desirable to obtain superior ride quality to that suitable for coal or grain.
One way to improve ride quality is to increase the dead sprung weight of the rail road car body. Another way to improve ride quality is to decrease the spring rate. Decreasing the spring rate involves further considerations. Historically the deck height of a flat car tended to be very closely related to the height of the upper flange of the center sill. This height was itself established by the height of the cap of the draft pocket. The size of the draft pocket was standardised on the basis of the coupler chosen, and the allowable heights for the coupler knuckle. The deck height usually worked out to about inches above top of rail. For some time auto rack cars were designed to a 19 ft height limit. To maximise the internal loading space, it has been considered desirable to lower the main deck as far as possible, particularly in tri-level cars. Since the lading is relatively light, the rail car trucks have tended to be light as well, such as 70 Ton trucks, as opposed to 100, 110 or 125 Ton trucks for coal, ore, or grain cars at 263,000, 286,000 or 315,000 lbs. gross weight on rail. Since the American Association of Railroads (AAR) specifies a minimum clearance of 5″ above the wheels, the combination of low deck height, deck clearance, and minimum wheel height set an effective upper limit on the spring travel, and reserve spring travel range available. If softer springs are used, the remaining room for spring travel below the decks may well not be sufficient to provide the desired reserve height. In consequence, the present inventor proposes, contrary to lowering the main deck, that the main deck be higher than 42 inches to allow for more spring travel.
As noted above, many previous auto rack cars have been built to a 19 ft height. Another major trend in recent years has been the advent of “double stack” intermodal container cars capable of carrying two shipping containers stacked one above the other in a well or to other freight cars falling within the 20 ft 2 in. height limit of AAR plate H. Many main lines have track clearance profiles that can accommodate double stack cars. Consequently, it is now possible to use auto rack cars built to the higher profile of the double stack intermodal container cars.
While decreasing the primary vertical bounce natural frequency appears to be advantageous for auto rack rail road cars generally, including single car unit auto rack rail road cars, articulated auto rack cars may also benefit not only from adding ballast, but from adding ballast preferentially to the end units near the coupler end trucks. As explained more fully in the description below, the interior trucks of articulated cars tend to be more heavily burdened than the end trucks, primarily because the interior trucks share loads from two adjacent car units, while the coupler end trucks only carry loads from one end of one car unit. It would be advantageous to even out this loading so that the trucks have roughly similar vertical bounce frequencies.
Three piece trucks currently in use tend to use friction dampers, sometimes assisted by hydraulic dampers such as can be mounted, for example, in the spring set. Friction damping has most typically been provided by using spring loaded blocks, or snubbers, mounted with the spring set, with the friction surface bearing against a mating friction surface of the columns of the side frames, or, if the snubber is mounted to the side frame, then the friction surface is mounted on the face of the truck bolster. There are a number of ways to do this. In some instances, as shown at p. 847 of the 1984 Car & Locomotive Cyclopedia lateral springs are housed in the end of the truck bolster, the lateral springs pushing horizontally outward on steel shoes that bear on the vertical faces of the side columns of the side frames. This provides roughly constant friction (subject to the wear of the friction faces), without regard to the degree of compression of the main springs of the suspension.
In another approach, as shown at p. 715 of the 1997 Car & Locomotive Cyclopedia, one of the forward springs in the main spring group, and one of the rearward springs in the main spring group bear upon the underside, or short side, of a wedge. One of the long sides, typically an hypotenuse of a wedge, engages a notch, or seat, formed near the outboard end of the truck bolster, and the third side has the friction face that abuts, and bears against, the friction face of the side column (either front or rear, as the case may be), of the side frame. The action of this pair of wedges then provides damping of the various truck motions. In this type of truck the friction force varies directly with the compression of the springs, and increases and decreases as the truck flexes. In the vertical bounce condition, both friction surfaces work in the same direction. In the warping direction (when one wheel rises or falls relative to the other wheel on the same side, thus causing the side frame to pivot about the truck bolster) the friction wedges work in opposite directions against the restoring force of the springs.
The “hunting” phenomenon has been noted above. Hunting generally occurs on tangent (i.e., straight) track as railcar speed increases. It is desirable for the hunting threshold to occur at a speed that is above the operating speed range of the rail car. During hunting the side frames tend to want to rotate about a vertical axis, to a non-perpendicular angular orientation relative to the truck bolster sometimes called “parallelogramming” or lozenging. This will tend to cause angular deflection of the spring group, and will tend to generate a squeezing force on opposite diagonal sides of the wedges, causing them to tend to bear against the side frame columns. This diagonal action will tend to generate a restoring moment working against the angular deflection. The moment arm of this restoring force is proportional to half the width of the wedge, since half of the friction plate lies to either side of the centreline of the side frame. This tends to be a relatively weak moment connection, and the wedge, even if wider than normal, tends to be positioned over a single spring in the spring group.
Typically, for a truck of fixed wheelbase length, there is a trade-off between wheel load equalisation and resistance to hunting. Where a car is used for carrying high density commodities at low speeds, there may tend to be a higher emphasis on maintaining wheel load equalisation. Where a car is light, and operates at high speed there will be a greater emphasis on avoiding hunting. In general, the parallelogram deformation of the truck in hunting is deterred by making the truck laterally more stiff. One approach to discouraging hunting is to use a transom, typically in the form of a channel running from between the side frames below the spring baskets. Another approach is to use a frame brace.
One way to address the hunting issue is to employ a truck having a longer wheelbase, or one whose length is proportionately great relative to its width. For example, at present two axle truck wheelbases may range from about 5′-3″ to 6′-0″. However, the standard North America track gauge is 4′-8½″, giving a wheelbase to track width ratio possibly as small as 1.12. At 6′-0″ the ratio is roughly 1.27. It would be preferable to employ a wheelbase having a longer aspect ratio relative to the track gauge. As described herein, one aspect of the present invention employs a truck with a longer wheelbase, preferably about 80 or 86 inches, giving a ratio of 1.42 or 1.52. This increase in wheelbase length may tend also to be benign in terms of wheel loading equalisation.
In a typical spring seat and spring group arrangement, the side frame window may typically be of the order of 21 inches in height from the spring seat base to the underside of the overarching compression member, and the width of the side frame window between the wear plates on the side frame columns is typically about 18″, giving a side frame window that is taller than wide in the ratio of about 7:6. Similarly, the bottom spring seat has a base that is typically about 18 inches long to correspond to the width of the side frame window, and about 16 inches wide in the transverse direction, that is being longer than wide. It may be advantageous to make the side frame windows wider, and the spring seat correspondingly longer to accommodate larger diameter long travel springs with a softer spring rate. At the same time, lengthening the wheel base of the truck may also be advantageous since it is thought that a longer wheelbase may ameliorate truck hunting performance, as noted above. Such a design change is counter-intuitive since it may generally be desired to keep truck size small, and widening the unsupported window span may not have been considered desirable heretofore.
Another way to raise the hunting threshold is to increase the parallelogram stiffness between the bolster and the side frames. It is possible, as described herein, to employ pairs of wedges, of comparable size to those previously used, the two wedges being placed side by side and each individually supported by a different spring, or being the outer two wedges in a three deep spring group, to give a larger moment arm to the restoring force and to the damping associated with that force.
The use of multiple variable friction force dampers in which the wedges are mounted over members of the spring group, is shown in U.S. Pat. No. 3,714,905 of Barber, issued Feb. 6, 1973. The damper arrangement shown by Barber is not apparently presently available in the market, and does not seem ever to have been made available commercially.
Notably, the damper wedges shown in Barber appear to have relatively sharply angled wedges, with an included angle between the friction face (i.e., the face bearing against the side frame column) and the sliding face (i.e., the angled face seated in the damper pocket formed in the bolster, typically the hypotenuse) of roughly 35 degrees. The angle of the third, or opposite, horizontal side face, namely the face that seats on top of the vertically oriented spring, is the complementary angle, in this example, being about 55 degrees. It should be noted that as the angle of the wedge becomes more acute, (i.e., decreasing from about 35 degrees) the wedge may have an undesirable tendency to jam in the pocket, rather than slide.
Barber, above, shows a spring group of variously sized coils with four relatively small corner coils loading the four relatively sharp angled dampers. From the relative sizes of the springs illustrated, it appears that Barber was contemplating a spring group of relatively traditional capacity—a load of about 80,000 lbs., at a “solid” condition of 3 1/16 inches of travel, for example, and an overall spring rate for the group of about 25,000 lbs/inch, to give 2 inches of overall rail car static deflection for about 200,000 lbs live load.
Apparently keeping roughly the same relative amount of damping overall as for a single damper, Barber appears to employ individual B331 coils (k=538 lb/in, (+/−)) under each friction damper, rather than a B432 coil (k=1030 lb/in, (+/−)) as might typically have been used under a single damper for a spring group of the same capacity. As such, it appears that Barber contemplated that springs accounting for somewhat less than 15% of the overall spring group stiffness would underlie the dampers.
These spring stiffnesses might typically be suitable for a rail road car carrying iron ore, grain or coal, where the lading is not overly fragile, and the design ratio of live load to dead sprung load is typically greater than 3:1. It might not be advantageous for a rail road car for transporting automobiles, auto parts, consumer electronics or other white goods of relatively low density and high value where the design ratio of live load to dead sprung load may be well less than 2:1, and quite possibly lying in the range of 0.4:1 to 1:1.
It has been noted that the frictional force produced by friction damper wedges differs depending on whether the damper is being loaded, or unloaded. In the terminology employed, the damper is being “loaded” when the bolster is moving downward in the sideframe window, since the spring force is increasing, and hence the load, or force on the damper is increasing. Similarly, the damper is being “unloaded” when the bolster is moving upward toward the top of the sideframe window, since the force in the springs, and hence the load in the wedges, is decreasing.
The equations can be written as
            While      ⁢                          ⁢      loading      ⁢                          ⁢              :            ⁢                          ⁢              F        d              =                  μ        c            ⁢              F        s            ⁢                        (                                    Cot              ⁢                              (                Φ                )                                      -                          μ              s                                )                                      1            +                                          (                                                      μ                    s                                    -                                      μ                    c                                                  )                            ⁢                              Cot                ⁡                                  (                  Φ                  )                                                      +                                          μ                s                            ⁢                              μ                c                                              )                                While      ⁢                          ⁢      unloading      ⁢                          ⁢              :            ⁢                          ⁢              F        d              =                  μ        c            ⁢              F        s            ⁢                        (                                    Cot              ⁢                              (                Φ                )                                      +                          μ              s                                )                                      1            +                                          (                                                      μ                    s                                    -                                      μ                    c                                                  )                            ⁢                              Cot                ⁡                                  (                  Φ                  )                                                      +                                          μ                s                            ⁢                              μ                c                                              )                    Where:                Fd=friction force on the sideframe column        Fs=force in the spring        μs=friction coefficient of the angled face on the bolster        μc is the coefficient of friction against the sideframe column        Φ is the included angle between the angled face on the bolster and the friction face bearing against the column        
For a given angle, a friction load factor, Cf can be determined as Cf=Fd/Fs. This load factor Cf will tend to be different depending on whether the bolster is moving up or down. A graph of upward and downward load factors as a function of wedge angle is shown in FIG. 7 based on a μs of 0.2 and a μc of 0.4, values which are thought to be roughly representative of service conditions.
When the wheels encounter a perturbation in the rail, their reaction to the perturbation will tend to transmit a force through the suspension into the rail road car body. The force transmitted will tend to be the sum of the spring force plus the friction force in the dampers. For a relatively gentle ride, it is desirable that the damping force as the wheels move up relative to the car body not be excessive, and that the damping be stronger when the car body is moving upward relative to the wheels.
With a relatively sharply angled wedge, as typified by wedges in the 30-35 degree range such as appear to be shown by Barber, and as employed in wedges known to be commonly in use, the load factor may tend to be significantly higher when the bolster is moving downward relative to the side frame than when the bolster is moving upward. It may be desirable to lessen, or reverse this relationship, as may tend to occur for angles above about 40 to 45 degrees. (See FIG. 7).
In the past, spring groups have been arranged such that the spring loading under the dampers has been proportionately small. That is, the dampers have typically been seated on side spring coils, as shown in the AAR standard spring groupings shown in the 1997 Car & Locomotive Cyclopedia at pages 743-746, in which the side spring coils, inner and outer as may be, are often B321, B331, B421, B422, B432, or B433 springs as compared to the main spring coils, such that the springs under the dampers have lower spring rates than the other coil combinations in the other positions in the spring group. As such, the dampers may be driven by less than 15% of the total spring stiffness of the group generally.
In U.S. Pat. No. 5,046,431 of Wagner, issued Sep. 10, 1991, the standard inboard-and-outboard gib arrangement on the truck bolster was replaced by a single central gib mounted on the side frame column for engaging the shoulders of a vertical channel defined in the end of the truck bolster. In doing this, the damper was split into inboard and outboard portions, and, further, the inboard and outboard portions, rather than lying in a common transverse vertical plane, were angled in an outwardly splayed orientation.
Wagner's gib and damper arrangement may not necessarily be desirable in obtaining a desired level of ride quality. In obtaining a soft ride it may be desirable that the truck be relatively soft not only in the vertical bounce direction, but also in the transverse direction, such that lateral track perturbations can be taken up in the suspension, rather than be transmitted to the car body, (and hence to the lading), as may tend undesirably to happen when the gibs bottom out (i.e., come into hard abutting contact with the side frame) at the limit of horizontal travel.
The present inventor has found it desirable that there be an allowance for lateral travel of the truck bolster relative to the wheels of the order of 1 to 1½ inches to either side of a neutral central position. Wagner does not appear to have been concerned with this issue. On the contrary, Wagner appears to show quite a tight gib clearance, with relatively little travel before solid contact. Furthermore, transverse displacement of the truck bolster relative to the side frame is typically resiliently resisted by the horizontal shear in the spring groups, and by the pendulum motion of the side frames rocking on the crowns of the bearing adapters, these two components being combined like springs in series. Wagner's canted dampers appear to make lateral translation of the bolster stiffer, rather than softer. This may not be advantageous for relatively fragile lading. In the view of the present inventor, while it is advantageous to increase resistance to the hunting phenomenon, it may not be advantageous to do so at the expense of increasing lateral stiffness.
It is desirable that a relatively larger portion of the spring effort be used to load the dampers, with the employment of a larger damper wedge angle. As such, the same magnitude of damping force may tend to be achieved with a combination of relatively softer springs than previously used, with a larger included angle in the wedges. Alternatively, a greater damping force than before may be achieved with wedges having a relatively modest angle with springs of the same stiffness as before, the included angle being chosen in the 45 to 65 degree range. The opportunity to vary wedge angle and spring stiffness thus gives an opportunity to tune the amount of damping in some measure. In addition, it would be advantageous to use a larger included angle in the wedge, both for these reasons, and because wedges with a larger included angle may tend to be less prone to jamming and may result in more favourable dynamic behaviour as indicated by FIG. 7.
In the damper groups themselves, it is thought that parallelogram deflection of the truck such that the truck bolster is not perpendicular to the side frame, as during hunting, may tend to cause the dampers to try to twist angularly in the damper seats. In that situation one corner of the damper may tend to be squeezed more tightly than the other. As a result, the tighter corner may try to retract relative to the less tight corner, causing the damper wedge to squirm and rotate somewhat in the pocket. This tendency to twist may also tend to reduce the squaring, or restoring force that tends to move the truck back into a condition in which the truck bolster is square relative to the side frames.
Consequently, it may be desirable to discourage this twisting motion by limiting the freedom to twist, as, for example, by introducing a groove or ridge, or keyway, or channel feature to govern the operation of the spring in the damper pocket. It may also be advantageous to use a split wedge to discourage twisting, such that one portion of the wedge can move relative to the other, thus finding a different position in a linear sense without necessarily forcing the other portion to twist. Further still, it may be advantageous to employ a means for encouraging a laterally inboard portion of the damper, or damper group, to be biased to its most laterally inboard position, and a laterally outboard portion of the damper, or the damper group, to be biased to its most laterally outboard position. In that way, the moment arm of the restoring force may tend to remain closer to its largest value. One way to do this, as described in the description of the invention, below, is to add a secondary angle to the wedge.
In the terminology herein, wedges have a primary angle Φ, namely the included angle between (a) the sloped damper pocket face mounted to the truck bolster, and (b) the side frame column face, as seen looking from the end of the bolster toward the truck center. This is the included angle described above. A secondary angle is defined in the plane of angle Φ, namely a plane perpendicular to the vertical longitudinal plane of the (undeflected) side frame, tilted from the vertical at the primary angle. That is, this plane is parallel to the (undeflected) long axis of the truck bolster, and taken as if sighting along the back side (hypotenuse) of the damper.
The secondary angle β is defined as the lateral rake angle seen when looking at the damper parallel to the plane of angle Φ. As the suspension works in response to track perturbations, the wedge forces acting on the secondary angle will tend to urge the damper either inboard or outboard according to the angle chosen. Inasmuch as the tapered region of the wedge may be quite thin in terms of vertical through-thickness, it may be desirable to step the sliding face of the wedge (and the co-operating face of the bolster seat) into two or more portions. This may be particularly so if the angle of the wedge is large.
Split wedges and two part wedges having a chevron, or chevron like, profile when seen in the view of the secondary angle can be used. Historically, split wedges have been deployed as a pair over a single spring, the split tending to permit the wedges to seat better, and to remain better seated, under twisting condition than might otherwise be the case. The chevron profile of a solid wedge may tend to have the same intent of preventing rotation of the sliding face of the wedge relative to the bolster in the plane of the primary angle of the wedge. Split wedges and compound profile wedges can be employed in pairs as described herein.
In a further variation, where a single broad wedge is used, with a compound or other profile, it may be desirable to seat the wedge on two or more springs in an inboard-and-outboard orientation to create a restoring moment such as might not tend to be achieved by a single spring alone. That is, even if a single large wedge is used, the use of two, spaced apart springs may tend to generate a restoring moment if the wedge tries to twist, since the deflection of one spring may then be greater that the other.
When the dampers are placed in pairs, either immediately side-by-side or with spacing between the pairs, the restoring moment for squaring the truck will tend not only to be due to the increase in compression to one set of springs due to the extra tendency to squeeze the dampers downward in the pocket, but due to the difference in compression between the springs that react to the extra squeezing of one diagonal set of dampers and the springs that act against the opposite diagonal pair that will tend to be less tightly squeezed.
The bolster is typically permitted to travel laterally to either side relative to the side frames, and for the side frames to have limited angular rotation about an axis parallel to the longitudinal axis of the rail car more generally. It is desirable that after an initial perturbation, the bolster return to a central, angularly squared position. An increase in the normal force at the friction face, as discussed, may tend to return the side frames to a “square” condition relative to the truck bolster. In sideways displacement, return of the truck to a centered position may tend to cease when the friction in the dampers matches the lateral restoring force in the spring groups. This tendency may be reduced by the tendency of the springs to return to a laterally centered position as the truck works in the vertical bounce and warp conditions. However, it may be desirable to enhance this restoring tendency. In the view of the present inventor it may be advantageous to install some, or all of the springs in the inner and outer rows of the spring group at a slight anhedral angle relative to each other, so that they form a symmetrical V.